(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Biodiversity Heritage Library | Children's Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "A dipstick ELISA for rapid detection of human blood meals in mosquitoes."

16 



Journal of the American Mosquito Control Association 



Vol. 7, No. 1 



A DIPSTICK ELISA FOR RAPID DETECTION OF HUMAN BLOOD 

MEALS IN MOSQUITOES 1 

H. M. SAVAGE, 2 J. F. DUNCAN, D. R. ROBERTS and L. L. SHOLDT 

Department of Preventive Medicine and Biometrics, Uniformed Services University of the Health Sciences, 
4301 Jones Bridge Road, Bethesda, MD 20814-4799 

ABSTRACT. A dot-enzyme-linked immunosorbent assay, dot-ELISA, that allows for identification 
of human blood meals in mosquitoes in less than 2 h is presented. Strips of mylar-backed nitrocellulose 
paper, the dipstick, may be inoculated with anti-human capture antibody, blocked, dried and stored for 
at least 3 months before use. The Dipstick ELISA consistently detected human blood meals at 24 h post- 
blood meal in frozen Anopheles mosquitoes and at 32 h post-blood meal on samples eluted off filter- 
paper smears. This ELISA detects human IgG at dilutions of 1:25,600, and produces strong positive 
results at dilutions between 1:400 and 1:12,800. The Dipstick ELISA is highly specific, and no false 
positives were detected when tested against cow, horse, goat, dog, cat, pig, rabbit, mouse, rat, chicken, 
raccoon and opossum sera. All reagents for the assay are commercially available. The Dipstick ELISA 
meets requirements for a rapid and simple assay for the identification of human blood meals and should 
have particular application to short-term field studies and emergency epidemiological investigations. A 
modified protocol, the Trough ELISA, which treats dipsticks jointly in disposable pipet troughs during 
the conjugate and substrate steps, was developed and produced as a kit. The Trough ELISA produces 1- 
4% random false positives, and we recommend that the Trough ELISA not be used. 



INTRODUCTION 

Identification of blood meals in hematopha- 
gous arthropods is an important tool in epide- 
miological investigations of vector-borne dis- 
eases. In particular, an assessment of the pro- 
portion of feeds taken on man, or the Human 
Blood Meal Index (HBI), is an essential variable 
in epidemiological investigations of vector- 
borne diseases such as malaria, and is needed to 
assess the vectorial capacity of each vector pop- 
ulation. A number of biochemical and immuno- 
logical techniques have been developed and suc- 
cessfully used to identify blood meal sources in 
mosquitoes (Washino and Tempelis 1983). 
Among these methods, enzyme-linked immu- 
nosorbent assay (ELISA) techniques have been 
demonstrated to provide both high sensitivity 
and specificity (Anonymous 1987). The World 
Health Organization (1987) and others recog- 
nized the demand for a rapid, accurate blood 
meal identification assay that could be con- 
ducted under field conditions or in simply 
equipped laboratories. A simple, rapid assay 
could provide entomologists and epidemiologists 
with real time data, and circumvent the major 
criticism of the standard procedure of shipping 
filter smears to a central processing center 
(World Health Organization 1987). 



1 The views of the authors do not purport to reflect 
the positions of the U.S. Department of Defense. The 
mention of proprietary products does not constitute 
endorsement by the U.S. Department of Defense. 

2 Current address: Medical Entomology and Ecology 
Branch, DVBID, CDC, P.O. Box 2087, Ft. Collins, CO 
80522-2087. 



Herein, we present a new dot-ELISA method 
for the identification of human blood meals in 
mosquitoes that can be completed in less than 2 
h when using preblocked dipsticks. This tech- 
nique is an indirect or sandwich type ELISA 
that is performed on a mylar-backed nitrocellu- 
lose paper which can be cut into strips to be 
used as dipsticks. All reagents needed to conduct 
this ELISA are commercially available, and both 
the capture and conjugated antibodies are affin- 
ity purified, which reduces cross-reactivity and 
limits the need to adsorb antiserum with cross- 
reacting blood sera. 

Based upon initial blind testing of 14 different 
blood meal assays, the ELISA presented below 
was selected by the World Health Organization 
for further testing and development. Selection 
criteria included sensitivity, specificity, simplic- 
ity and minimum-time required to conduct the 
test under field conditions and/or in simply 
equipped laboratories located in disease endemic 
areas (V. Houba, personal communication, 1987; 
World Health Organization 1987). 

Unlike other ELISAs used in blood meal 
analysis (Beier et al. 1988, Lombardi and Espos- 
ito 1986, Service et al. 1986), the dipsticks pro- 
vide a permanent record of test results. This 
Dipstick ELISA meets requirements for a fast 
and simple assay for the identification of human 
blood meals when testing a relatively small num- 
ber of specimens, and should have particular 
application to field studies and to short-term or 
emergency epidemiological investigations. 

MATERIALS AND METHODS 

Dipstick ELISA protocol: The basic protocol 
is outlined below and presented schematically 
in Fig. 1. All procedures are conducted at room 



March 1991 



Dipstick ELISA for Blood Meal Identification 



17 



temperature. Dipsticks must be processed indi- 
vidually throughout the procedure. Dipsticks 
may be exposed to antigen, conjugate and sub- 
strate solutions in separate 1.5-ml microcentri- 
fuge tubes (Eppendorf #2236380-8, Thomas Sci- 
entific, Philadelphia, PA) or in wells of styrene 
microtiter plates {for example, Dynatech Im- 
mulon 2, Dynatech Inc., Alexandria, VA). How- 
ever, polyvinyl chloride microtiter plates are 
unsuitable. 

1) Marking the mylar-backed nitrocellulose 
(NC) sheets. The mylar-backed 0.45-^m nitro- 
cellulose membrane (BA85) is supplied in 15 X 
15 cm sheets (Schleicher & Schuell Inc., Keene, 
NH). The use of plastic gloves when handling 
nitrocellulose is recommended to keep body oils 
off the membrane. However, normal handling of 
the membrane with clean hands will not affect 
the results of the assay. The nitrocellulose sur- 
face of the sheets may be marked with a high 
quality ballpoint pen and ruler to produce dip- 
sticks of the desired size. We prefer dipsticks 
that are approximately 3.75 cm long, and 0.4- 



0.5 cm wide with a line drawn 0.5 cm from one 
end to produce a small box. Sheets should be 
marked such that the small boxes (the site where 
the capture antibody will be placed) are not 
located along the sheet margins. These dimen- 
sions allow for 4 rows of dipsticks per sheet, or 
about 148 dipsticks per sheet. The desired num- 
ber of dipsticks may then be cut from the sheet 
as a single unit or block (not individually). Dip- 
sticks may be numbered immediately if your 
testing needs are known, or the dipsticks may 
be numbered as they are needed after blocking. 
Only number dipsticks when they are thor- 
oughly dry. 

2) Preparation of stock capture antibody so- 
lution. Reconstitute a 1.0-mg vial of lyophilized, 
unlabeled, goat anti-human IgG (H + L) (Kir- 
kegaard & Perry Lab. Inc., Gaithersburg, MD) 
with 1.0 ml of a 50% glycerol solution. The 
reconstituted capture antibody remains stable 
for at least 3 months at 4-8°C; however, aliquots 
of the stock capture antibody solution may be 
stored at -20°C for 1 year. 




WET NC MEMBRANE 
WITH WATER 

15MIN. 



BLOCKING 
BUFFER 

30 MIN. 



ADD CAPTURE ANTIBODY, 1 ul spot 



O 





4 

UNKNOWN 
ANTIGEN 




45 MIN. 



CONJUGATED 
ANTIBODY 

25 MIN. 



SUBSTRATE 
(BCIP/NBT) 

8-15 MIN. 



WASH 




O 



POSITIVE 



NEGATIVE 



an?Metho^rdetaiL StePS ^ DiPSUCk ^ Pr ° t0C01 W ' th P ° sitiTC md "*«™ **k, See Materials 



18 



Journal of the American Mosquito Control Association 



Vol. 7, No. 1 



3) Wetting the nitrocellulose membrane. Cut 
a section from the marked sheet that includes 
the desired number of dipsticks and wet, or soak, 
in distilled water for 15 min to prepare the 
membrane for the adsorption of capture anti- 
body. After 15 min, remove the membrane from 
the water and lightly blot with a clean paper 
towel to remove excess moisture. 

4) Adsorption of capture antibody. The work- 
ing capture antibody solution is a 0.2 mg/ml 
solution, or a 1:5 dilution of the reconstituted 
capture antibody in phosphate saline buffer 
(PBS), pH 8.0-8.3. The working solution should 
be made immediately before use. A small num- 
ber of dipsticks, 38 or less, should be treated at 
one time because the dipsticks must be prepared 
immediately after blotting in the wetting step 
above. After blotting, but before the nitrocellu- 
lose membrane becomes dry, place a 1.0-/il drop 
of the capture antibody solution in the middle 
of the small boxed end of each dipstick. The 
capture antibody will bind and dry in less than 
60 sec. Capture antibody may be applied with 
disposable 5-//1 microcapillary displacement pi- 
pettes (Cole-Parmer Instrument Co., Chicago, 
IL), or a variety of repeating pipettes. Allow 
dipsticks to dry for 2-10 min. 

5) Blocking. Submerge dipsticks in a blocking 
solution [a 1:5 dilution of a 2% milk diluent/ 
blocking concentrate (Kirkegaard & Perry Lab. 
Inc.) in PBS] for 30 min. Remove dipsticks and 
air dry. Dipsticks may be used immediately or 
stored desiccated in a refrigerator or at room 
temperature (less than 22° C) for 90 days. Al- 
though blocking is not absolutely necessary, it 
reduces background color and also serves to 
stabilize the capture antibody if dipsticks are to 
be stored. 

6) Preparation of mosquito samples and con- 
trols. Whole mosquito specimens or mosquito 
abdomens may be prepared by triturating an 
individual specimen in 0.5 ml of PBS in a 1.5- 
ml microcentrifuge tube with a plastic pestle 
(Kontes Scientific Glassware/Instruments, 
Vineland, NJ). Freshly blooded mosquitoes 
should be held at room temperature for at least 
2 h before processing or smeared onto filter 
paper and punches of filter paper used to prepare 
the antigen source. When using filter-paper 
smears as the antigen source place one or 2 
punches of the filter paper in a microcentrifuge 
tube with 300 fi\ of a diluent solution (1:20 
dilution of a 2% milk diluent/blocking concen- 
trate in PBS). Allow the solution to stand at 
room temperature for 2 h before testing. For 
each test, at least one negative and one positive 
control must be run. A 1:201 dilution of serum 
in PBS from one or more nonhuman sources 
should be used as negative controls, and a 1:201 
to 1:401 dilution of human serum as the positive 



control. Alternatively, mosquitoes blooded on 
known hosts could be used as controls. Male 
mosquitoes should not be used as negative con- 
trols in this or any other blood meal ELISA, 
except as reagent controls. 

7) Antigen. Label or number each dipstick for 
permanent association with a mosquito speci- 
men or control. Cut the desired number of dip- 
sticks, usually 24 per test, from a treated nitro- 
cellulose sheet. Place the treated end of each 
dipstick in a mosquito homogenate or antigen 
source for 45 min. Dipsticks may be placed 
directly into the microcentrifuge tubes from step 
6 or into plate-wells containing 200-250 fi\ of 
antigen source. 

8) Washing and blotting. Remove dipsticks 
from antigen source and wash by holding them 
individually under flowing distilled water or 
clean tap water for 10-30 sec. Do not place 
dipsticks together in a tube for washing. Shake 
or blot excess water from dipsticks. 

9) Preparation of enzyme-conjugated anti- 
body solutions. The stock conjugate solution is 
prepared by reconstituting a 0.1-mg vial of ly- 
ophilized, alkaline phosphatase conjugated, goat 
anti-human IgG (H -I- L) (Kirkegaard & Perry 
Lab. Inc.) and storing as described for the cap- 
ture antibody in step 2. The working conjugate 
solution is a 0.2-jug/ml solution, or a 1:501 dilu- 
tion of the stock conjugated antibody in a di- 
luent solution [diluent solution = 1:20 dilution 
of a 2% milk diluent/blocking concentrate in 
PBS], and should be prepared immediately be- 
fore use. 

10) Conjugate step. Place dipsticks individ- 
ually into microtiter plate-wells or tubes with 
200-250 fA of working conjugate solution for 20- 
25 min. 

11) Wash and blot as in step 8. 

12) Substrate. Place dipsticks individually 
into wells containing 200-250 fd of the 3 part 
BCIP/NBT (5-bromo-4-chloro-3-indolyl phos- 
phate and nitroblue tetrazolium) substrate so- 
lution (Kirkegaard & Perry Lab. Inc.). Place the 
required amount of Tris buffer from the sub- 
strate kit into a test tube before use so that 
the buffer may warm to room temperature. The 
BCIP and NBT concentrate solutions should be 
added to the Tris buffer immediately before use. 
A small purple dot will develop in 8-15 min on 
positive dipsticks. Reaction time should be 
judged by inspection of the positive control dip- 
stick. All dipsticks should be removed simulta- 
neously when a purple dot is easily visible on 
the positive control dipstick. As a general rule, 
it is better to remove the dipsticks a little too 
early rather than a little too late, because with 
excessive exposure, all dipsticks that have been 
exposed to animal IgG will develop to some 
extent. Air dry dipsticks and attach to a sheet 



March 1991 



Dipstick ELISA for Blood Meal Identification 



19 



of heavy paper or a collection form as a perma- 
nent record of your assay. 

13) Interpretation of results. Test results are 
interpreted by comparing the reaction on each 
unknown dipstick with the positive and negative 
control dipsticks. Test resolution improves after 
the dipsticks are thoroughly dried, 45-60 min, 
because background color fades upon drying 
which increases the contrast between positive 
spots and surrounding areas. 

Evaluation of Dipstick ELISA by World Health 
Organization (WHO)-interlaboratory trial: In 
1987, the above ELISA was evaluated by blind 
testing of mosquito smears and whole blood 
samples on filter paper in an interlaboratory 
trial, sponsored by WHO-Vector Biology Con- 
trol Division (WHO/VBC) and organized by 
WHO-Immunology Research and Training Cen- 
ter (WHO-IRTC), Geneva, Switzerland (World 
Health Organization 1987). The unknowns in- 
cluded filter-paper smears of Anopheles ste- 
phensi Liston killed at 0, 8, 24, 32 and 48 h after 
feeding, and a mosquito fed on glucose only. 
Whole blood, filter-paper samples were prepared 
by placing a 50-/d dot of each blood sample on 
a Whatman No. 1 filter paper and drying at 
37° C for 1 h before storage and shipment. Whole 
blood samples included: cow, dog, pig, goat, 
chicken, 2 pure human samples, one 1:10 mix- 
ture of human:cow, and one 1:100 mixture of 
humanxow. Samples were tested, and the re- 
sults and dipsticks were forwarded to WHO for 
evaluation. 

Final evaluation: The experiments described 
in this final evaluation are only those used to 
document the final version of the ELISA, and 
the testing of various parameters are not re- 
ported on herein. 

Specimens of An. albimanus Wied. were given 
their first blood meal on human volunteers, ex- 
amined for the presence of a blood meal and 
sorted. Blooded specimens were maintained in 
an insectary at 27 ± 1°C. After 24 h, mosquitoes 
were provided with a sugar solution. At 1 h and 
at 6 h post-blood meal, and every 6 h thereafter, 
9 specimens were killed and placed individually 
into 1.5-ml tubes, and stored at -70° C until 
analyzed by Dipstick ELISA. Specimens were 
analyzed by Dipstick ELISA with a 1:401 dilu- 
tion of human serum in PBS used as positive 
controls, and a 1:201 dilution of cow serum in 
PBS used as negative controls. 

To investigate specificity and sensitivity of 
the Dipstick ELISA, we conducted a series of 
ELISAs, each of which included a 12-step serial 
dilution of human serum in PBS starting at 
1:100 and running through 1:204,800 (12 dip- 
sticks), and 1:200 dilutions of the following an- 
imal sera in PBS (12 dipsticks): cow, horse, goat 



dog, cat, pig, rabbit, mouse, rat, chicken, raccoon 
and opossum. 

A modified version, the Trough ELISA, for 
field kits and its evaluation by WHO and the 
Uniformed Services University of the Health Sci- 
ences (USUHS): In 1988-89, we developed a 
modification of our basic dipstick protocol that 
employed disposable multichannel pipet troughs 
(Costar, Cambridge, MA) for joint treatment of 
dipsticks in the conjugate and substrate steps. 
We produced 7 kits, each of which included 1,000 
preinoculated and preblocked dipsticks and all 
items needed to conduct the Trough ELISA. 
These 7 kits were evaluated in a field trial, 
sponsored by WHO/VBC and organized by 
WHO-IRTC, Geneva, Switzerland, in collabo- 
ration with 5 institutions in malaria endemic 
areas (Brazil, Cameroon, India, Malaysia and 
Tanzania) and by WHO-IRTC, Geneva, Swit- 
zerland. A description of the kits with detailed 
instructions for the Trough ELISA prepared by 
us and included in each kit will be published by 
WHO (V. Houba, personal communication, 
1990) and will not be repeated herein. 

RESULTS AND DISCUSSION 

Evaluation of Dipstick ELISA by WHO-inter- 
laboratory trial: Filter-paper smears from the 
digestion series and a glucose-fed specimen were 
prepared and divided into 3 aliquots each. Dip- 
sticks from all 3 aliquots were positive for sam- 
ples from 0, 8, 24 and 32 h post-blood meal 
(World Health Organization 1987, Anonymous 
1987). The results from the 48-h post-bloodmeal 
sample were ambiguous, with 2 dipsticks scored 
as weak positives and one dipstick scored nega- 
tive. Tests of the glucose-fed mosquito resulted 
in 3 negatives. 

The Dipstick ELISA correctly identified the 
following blood samples as nonhuman: cow, dog, 
pig, goat and chicken. The assay correctly iden- 
tified the following blood samples as human: 2 
separate pure human samples, and a 1:10 mix- 
ture of human and cow blood. The single error 
observed during this blind evaluation was a fail- 
ure to identify the 1:100 mixture of humanrcow 
blood as positive for human (World Health Or- 
ganization 1987). For whole blood samples, the 
efficiency of the ELISA was 90%. [Efficiency = 
((true positives 4- true negatives)/(true positives 
+ false positives + true negatives + false nega- 
tives)) X 100]. 

Final evaluation and troubleshooting: Speci- 
mens of An. albimanus engorged on humans 
were tested at time intervals post-blood meal to 
investigate the sensitivity of the Dipstick ELISA 
with respect to blood meal digestion. Results 
indicate that the assay consistently produces 
strong positive reactions at digestion periods of 



20 


Journal 


OF THE 


American Mosquito Control Association 


Vol. 


7, No. 1 


Table 1. Sensitivity of Dipstick ELISA, see Materials and Methods section for protocol, 
after digestion by Anopheles albimanus at 27 ± 1°C. Specimens engorged on human s 


to human blood meal 
ubjects at time 0. 


Hours postblood meal 






Digestion series — replicate number 






1 


2 


3 


4 


5 


6 


7 


8 


9 


1 




++ 


++ 


++ 


-f+ 


++ 


++ 


++ 


++ 


++ 


6 




++ 


++ 


++ 


++ 


++ 


++ 


++ 


++ 


++ 


12 




++ 


++ 


++ 


++ 


++ 


++ 


++ 


++ 


++ 


18 




++ 


++ 


++ 


++ 


++ 


++ 


++ 


++ 


++ 


24 




+ 


± 


-f 


+ 


++ 


+ 


++ 


++ 


+ 


30 




± 


+ 


± 


± 


+ 


+ 


+ 


- 


+ 


36 




+ 


+ 


± 


+ 


+ 


++ 


+ 


± 


± 


42 
48 




— 


: 


+ 


— 


_ 


_ 


_ 


: 


: 


54 

Negative control 

Positive control 


++ 


++ 


++ 


++ 


++ 


++ 


++ 


++ 


++ 



++ = strong positive; + = positive, or definite positive but light; ± = ghost image of uncertain status, scored 
as negative; — = negative. 



18 h or less on frozen specimens (Table 1). At 
24 h post-blood meal, 8 of the 9 specimens gave 
positive to strong positive results, whereas one 
specimen produced a ghost image of uncertain 
status and was scored negative. Most dipsticks 
from tests conducted on specimens of 30 and 36 
h post-blood meal developed ghost images that 
were scored as negatives. At 42, 48 and 54 h 
post-blood meal, all dipsticks except one dis- 
played clear negative results. These results com- 
pare favorably with those of a direct, liquid- 
phase ELISA for blood meal identification 
(Beier et al. 1988). The liquid-phase ELISA 
(Beier et al. 1988) consistently detected human 
blood meals in frozen specimens up to 23 h post- 
blood meal with detectability falling off appre- 
ciably after 23 h. 

The sensitivity and specificity of the assay 
were documented by 8 separate replicates of the 
Dipstick ELISA, each of which included a 12- 
step serial dilution of human serum starting at 
a 1:100 dilution and running through 1:204,800, 
and 1:200 dilutions of 12 different animal sera 
(Table 2). Results indicate that the Dipstick 
ELISA is sensitive to human IgG at serum di- 
lutions of 1:25,600, and that the assay provides 
strong positive results at dilutions between 1:400 
and 1:12,800. Results from the 1:100 dilution 
indicate that very concentrated serum samples 
may yield inconsistent results with about 50% 
of dots appearing light enough to be considered 
strong ghost images, rather than positives. To 
reduce the likelihood of obtaining false negatives 
from newly engorged specimens we recommend 
(see step 6, Materials and Methods): that all 
freshly engorged specimens be held for 2 h before 
being processed; or that freshly engorged speci- 
mens be smeared onto filter paper and punches 
of the filter paper be used to prepare the antigen 
source. Additionally, human IgG in insect blood- 



meals appears to be better preserved by the 
filter-paper smear technique and by simple 
drying than by freezing. This effect is apparent 
in our Dipstick ELISA and also in the liquid- 
phase ELISA of Beier et al. (1988). In the Dip- 
stick ELISA, human blood meals in frozen spec- 
imens were detectable to 24 h post-blood meal, 
and human blood meals were consistently de- 
tected in smears of An. stephensi at 32 h post- 
blood meal (see Evaluation of Dipstick ELISA 
by WHO-interlaboratory trial). In the liquid- 
phase ELISA (Beier et al. 1988), blood meals 
from frozen specimens were detectable to 23 h 
post-blood meal, and human IgG was consist- 
ently detected at 32 h post-blood meal in speci- 
mens of An. gambiae s. lat. and Anopheles funes- 
tus Giles dried in a desiccator jar at room tem- 
perature. 

The Dipstick ELISA correctly identified all 
12 animal sera as nonhuman in all 8 tests. The 
distribution of ghost images among the 12 ani- 
mal sera appears random. It is unlikely that the 
ghost images represent true cross-reactivity as 
the animal species known to be most highly 
cross-reactive, e.g., dog, produced consistently 
negative results. However, in other modifica- 
tions of this ELISA we have observed limited 
cross- reactivity between the human antibody 
and cow, cat, dog, mouse and rat. These and 
other nontested host species may cross-react to 
produce strong ghost images. Ghost images in 
negative control dipsticks will result in reduced 
specificity of the human test because each test 
dipstick is compared with the positive and neg- 
ative controls to interpret test results. If cross- 
reactivity appears to be a problem in a study 
area, it may be reduced by adding sera of cross- 
reacting species to the anti-human conjugate 
solution at a 1:501 to a 1:1,001 dilution. 

Strong ghost images and even weak positive 



March 1991 



Dipstick ELISA for Blood Meal Identification 



21 



Table 2. Results of 8 replicates of the Dipstick ELISA, see Materials and Methods section for protocol 
conducted on serial dilutions of human serum and 1:200 dilutions of 12 animal sera. 



Dilution 



1:100 

1:200 

1:400 

1:800 

1:1,600 

1:3,200 

1:6,400 

1:12,800 

1:25,600 

1:51,200 

1:102,400 

1:204,800 

1:200 

1:200 

1:200 

1:200 

1:200 

1:200 

1:200 

1:200 

1:200 

1:200 

1:200 

1:200 



Antigen 



Replicate number 



Human 

Human 

Human 

Human 

Human 

Human 

Human 

Human 

Human 

Human 

Human 

Human 

Cow 

Horse 

Goat 

Dog 

Cat 

Pig 

Rabbit 

Mouse 

Rat 

Chicken 

Raccoon 

Opossum 



7 



+ 

+ 

++ 
++ 
++ 
++ 
++ 
++ 
++ 

+ 



+ 
+ 

++ 
++ 
++ 
++ 
++ 
++ 
++ 
+ 



++ 
++ 
++ 
++ 
++ 
++ 
++ 
++ 



+ 
+ 

++ 
++ 
++ 
++ 
++ 
++ 
++ 
+ 



++ 
++ 
++ 
++ 
++ 
++ 
++ 

+ 

+ 



++ 
++ 
++ 
++ 
++ 
++ 
++ 



+ 
+ 
+ 

++ 
++ 
++ 
++ 

+ 



+ 

++ 
++ 
++ 
++ 
++ 
++ 
++ 



as neVtive r ; - g = P negaIive + = ^^ " ^^ ^^ ^ Ught; ± = gh ° St image of uncertain statu «- scored 



results may appear on known negative dipsticks 
for reasons other than cross-reactivity. The 
most common reasons for poor results are in- 
adequate individual washing of dipsticks, joint 
washing of dipsticks, or allowing the dipsticks 
to come into direct contact with each other. 
Another common mistake is the use of too many 
dipsticks in a single test. This assay is designed 
for the rapid testing of a relatively small number 
of mosquitoes, 20-28, at a time. The processing 
of large numbers of dipsticks during a single test 
is difficult because dipsticks must be individ- 
ually handled and washed. We have obtained 
consistently good results when using 24 dip- 
sticks per test, which allows for 22 unknowns, 
one positive control and one negative control! 
Overexposure to substrate or incorrectly mixing 
the substrate solution may also lead to difficul- 
ties in interpreting test results. Time in sub- 
strate solution depends on the temperature and 
the age of the substrate solution. We typically 
allow the Tris buffer to come up to room tem- 
perature before use. However, the BCIP and 
NBT concentrate solutions should be added to 
the Tns buffer in the substrate tube only im- 
mediately before use. Dipsticks should be re- 
moved from the substrate solution simultane- 
ously when a purple circular spot appears on the 
positive control dipstick. In our experience, this 



occurs between 8-20 min and commonly from 
8-14 min. As a general rule, it is better to remove 
the dipsticks a little too early rather than a little 
too late, because with excessive exposure all 
dipsticks that have been exposed to animal IgG 
will develop to some extent. 

A modified version, the Trough ELISA, for 
field kits and its evaluation by WHO and 
USUHS: The field trial of the Trough ELISA 
by WHO and cooperating laboratories has been 
completed and the results will be published sep- 
arately by WHO personnel. 

As a means of comparison with the basic 
protocol outlined in the methods section (Table 
2), we conducted 8 separate Trough ELISAs 
employing the same antigen sources (Table 3). 
Four false positives (4.17%) were observed 
among the 96 nonhuman antigen sources in the 
Trough ELISA (Table 3), versus no false posi- 
tives in the basic dipstick protocol (Table 2) 
Additionally, 22 ghost images (22.9%) were ob- 
served among the 96 nonhuman antigen sources 
in the Trough ELISA, versus 5 ghost images 
(5.2%) in the basic protocol (Table 2). Ghost 
images in and of themselves are not a serious 
problem as they can typically be recognized as 
such and scored as negatives. However, among 
dipsticks exposed to human serum, note that 
ghost images (±) and positives (+) are more 



22 



Journal of the American Mosquito Control Association 



Vol. 7, No. 1 



Table 3. Results of 8 replicates of the Trough ELISA, a modified Dipstick ELISA that treats dipsticks jointly 
in troughs during the conjugate and substrate steps, conducted on serial dilutions of human serum and 1:200 

dilutions of 12 animal sera. 













Replicate number 








Dilution 


Antigen 






































1 


2 


3 


4 


5 


6 


7 


8 


1:100 


Human 


+ 


± 


+ 


+ 


± 


+ 


+ 


+ 


1:200 


Human 


+ 


+ 


+ 


+ 


+ 


+ 


+ 


+ 


1:400 


Human 


+ 


+ 


+ 


+ 


+ 


++ 


++ 


+ 


1:800 


Human 


+ 


+ 


+ 


+ 


+ 


++ 


++ 


++ 


1:1,600 


Human 


+ 


+ 


+ 


++ 


++ 


++ 


++ 


++ 


1:3,200 


Human 


++ 


+ 


++ 


++ 


++ 


-+-+ 


++ 


++ 


1:6,400 


Human 


++ 


++ 


++ 


++ 


++ 


++ 


++ 


++ 


1:12,800 


Human 


++ 


++ 


++ 


++ 


++ 


++ 


+ 


+ 


1:25,600 


Human 


+ 


++ 


++ 


+ 


+ 


± 


- 


+ 


1:51,200 


Human 


+ 


± 


++ 


± 


+ 


± 


- 


— 


1:102,400 


Human 


± 


± 


± 


+ 


+ 


— 


— 


— 


1:204,800 


Human 


+ 


± 


± 


± 


± 


± 


- 


— 


1:200 


Cow 


- 


- 


± 


+ 


— 


- 


- 


— 


1:200 


Horse 


± 


± 


- 


- 


- 


- 


- 


— 


1:200 


Goat 


- 


- 


- 


+ 


± 


- 


- 


— 


1:200 


Dog 


- 


_ 


± 


- 


± 


— 


— 


— 


1:200 


Cat 


— 


± 


+ 


— 


± 


+ 


— 


— 


1:200 


Pig 


± 


- 


- 


- 


- 


- 


- 


+ 


1:200 


Rabbit 


- 


- 


- 


- 


- 


- 


— 


— 


1:200 


Mouse 


- 


- 


- 


± 


- 


- 


- 


— 


1:200 


Rat 


- 


± 


- 


- 


± 


- 


- 


— 


1:200 


Chicken 


+ 


- 


- 


+ 


± 


- 


— 


— 


1:200 


Raccoon 


± 


+ 


- 


+ 


± 


- 


— 


— 


1:200 


Opossum 


± 


+ 


- 


— 


~ 


— 


— 


— 



++ = strong positive; + = positive, or definite positive but light; ± = ghost image of uncertain status, scored 
as negative; — = negative. 



common in the Trough ELISA (Table 3) than 
in the basic dipstick protocol (Table 2), and that 
strong positives (++) are less common in the 
Trough ELISA than in the basic protocol. In 
total, the contrast between positives (+ and ++) 
and ghost images is less distinct in the Trough 
ELISA than in the basic protocol, which in- 
creases the uncertainty of scoring for ghost im- 
ages and light positive dipsticks. 

We have experimented extensively with other 
parameters of the Trough ELISA during the 
past year. Originally, we believed that it was 
possible to remove excess antigen by washing 
and by cutting the time in conjugate to a mini- 
mum, and to develop a test that allowed joint 
treatment of dipsticks in the conjugate and sub- 
strate steps. The number of ghost images and 
false positives can be reduced to less than half 
the levels observed in Table 3 by extensive wash- 
ing and by reducing the time in conjugate to 15- 
20 min. However, it is not possible to consist- 
ently remove all the excess antigen by washing, 
and antigen from one dipstick with excess anti- 
gen may contaminate the conjugate solution and 
produce ghost images or false positives on 
nearby dipsticks. We documented the presence 
of antigen contaminate in used conjugate solu- 
tions by employing used conjugate solutions as 



the antigen source in subsequent basic protocol 
Dipstick ELISAs, while unused conjugate solu- 
tions were used as control antigen sources. In 
each test, which was repeated 5 times, there was 
sufficient antigen in the used conjugate solu- 
tions to produce false positives or ghost images 
on all exposed dipsticks, whereas an equal num- 
ber of dipsticks exposed to unused conjugate 
solutions prepared at the same time gave con- 
sistently negative results as would be expected. 
Obviously, there was a sufficient amount of free 
antigen contaminant in the used conjugate so- 
lution to produce false positives. Therefore, we 
recommend that the Trough version of the Dip- 
stick ELISA not be used. 

CONCLUSIONS 

After considerable effort, we have come to the 
conclusion that each dipstick must be kept sep- 
arate in all steps. If a microtiter plate-well is 
employed in the antigen, conjugate and sub- 
strate steps outlined in the basic protocol, then 
false positives will be eliminated, ambiguous 
ghost images will be rare and the Dipstick 
ELISA will produce consistently reliable results. 
The Dipstick ELISA meets requirements for a 
rapid (less than 2 h when using preblocked dip- 



March 1991 



Dipstick ELISA for Blood Meal Identification 



23 



sticks) and simple assay for the identification of 
human blood meals in a relatively small number 
of specimens (24 dipsticks per test). The Dip- 
stick ELISA should have particular application 
to discrete field studies and to short-term or 
emergency epidemiological investigations. In- 
vestigators wishing to conduct a large number 
of blood meal assays for human and other host 
species, or those wishing to establish a labora- 
tory for the processing of a large number of 
specimens on a routine basis would be advised 
to set up a liquid-phase ELISA (Beier et al. 
1988, Service et al. 1986). 

ACKNOWLEDGMENTS 

This research was funded in part by each of 
the following: by project DI-MOD of the Na- 
tional Aeronautics and Space Administration 
(NASA) through contract W16,306 to USUHS, 
entitled "Application of Remote Sensing and 
Predictive Modeling to Malaria Transmission 
and Vector Ecology"; by a grant from the World 
Health Organization to D. R. Roberts entitled 
"Dot-ELISA for Arthropod Blood Meal Identi- 
fication"; and by a faculty development grant 
from USUHS to L. L. Sholdt entitled "A Dip- 
stick, Dot-ELISA, for Rapid Identification of 
Insect Blood Meals". The authors thank V. 
Houba, WHO-IRTC, Geneva for organizing the 
evaluation of the kits by cooperating WHO in- 



stitutions, for evaluating the Trough ELISA in 
his laboratory, and for helpful comments on the 
Trough ELISA and kit instructions. We thank 
R. Andre, USUHS, R. A. Wirtz, Walter Reed 
Army Institute of Research, and V. Houba, 
WHO-IRTC for comments on a draft of this 
manuscript. 

REFERENCES CITED 

Anonymous. 1987. Blood meal identification in vec- 
tors. Parasitol. Today 3:324-326. 

Beier, J. C, P. V. Perkins, R. A. Wirtz, J. Koros, D. 
Diggs, T. P. Gargan II and D. K. Koech. 1988. Blood 
meal identification by direct enzyme-linked immu- 
nosorbent assay (ELISA), tested on Anopheles (Dip- 
tera: Culicidae) in Kenya. J. Med. Entomol. 25: 
9-16. 

Lombardi, S. and F. Esposito. 1986. A new method for 
identification of the animal origin of mosquito blood 
meals by the immunobinding of peroxidase -anti- 
peroxidase complexes on nitrocellulose. J. Immunol. 
Meth. 86:1-5. 

Service, M. W., A. Voller and D. E. Bidwell. 1986. The 
enzyme-linked immunosorbent assay (ELISA) test 
for the identification of blood meals of haematopha- 
gous insects. Bull. Entomol. Res. 76:321-330. 

Washino, R. K. and C. H. Tempelis. 1983. Mosquito 
host blood meal identification: methodology and 
data analysis. Annu. Rev. Entomol. 28:179-201. 

World Health Organization. 1987. Development of 
new techniques for the identification of host blood 
meals in arthropod vectors, report of the interlabo- 
ratory trial 1986-1987. WHO/VBC/87.947:l-12.